Multiple Object Tracking on BDD100K val

Metric: mMOTA (higher is better)

LeaderboardDataset

Loading chart...

Results

Submit a result

Sort:

#	Model↕	mMOTA▼	Extra Data	Paper	Date↕	Code
1	ByteTrack	45.5	No	ByteTrack: Multi-Object Tracking by Associating ...	2021-10-13	Code
2	UNINEXT-H	44.2	No	Universal Instance Perception as Object Discover...	2023-03-12	Code
3	MOTRv2	43.6	No	MOTRv2: Bootstrapping End-to-End Multi-Object Tr...	2022-11-17	Code
4	QDTrack	42.1	No	QDTrack: Quasi-Dense Similarity Learning for App...	2022-10-12	Code
5	ContrasTR	41.7	No	Contrastive Learning for Multi-Object Tracking w...	2023-11-14	Code
6	Unicorn	41.2	No	Towards Grand Unification of Object Tracking	2022-07-14	Code
7	TETer	39.1	No	Tracking Every Thing in the Wild	2022-07-26	Code
8	QDTrack	36.6	No	Quasi-Dense Similarity Learning for Multiple Obj...	2020-06-11	Code
9	Adaptive-searching -windows Tracker	34.4	No	-	-	-

#1ByteTrackSOTA
45.5
mMOTA· 2021-10-13
ByteTrack: Multi-Object Tracking by Associating Every Detection Box Code
#2UNINEXT-H
44.2
mMOTA· 2023-03-12
Universal Instance Perception as Object Discovery and Retrieval Code
#3MOTRv2
43.6
mMOTA· 2022-11-17
MOTRv2: Bootstrapping End-to-End Multi-Object Tracking by Pretrained Object Detectors Code
#4QDTrack
42.1
mMOTA· 2022-10-12
QDTrack: Quasi-Dense Similarity Learning for Appearance-Only Multiple Object Tracking Code
#5ContrasTR
41.7
mMOTA· 2023-11-14
Contrastive Learning for Multi-Object Tracking with Transformers Code
#6Unicorn
41.2
mMOTA· 2022-07-14
Towards Grand Unification of Object Tracking Code
#7TETer
39.1
mMOTA· 2022-07-26
Tracking Every Thing in the Wild Code
#8QDTrackSOTA
36.6
mMOTA· 2020-06-11
Quasi-Dense Similarity Learning for Multiple Object Tracking Code
#9Adaptive-searching -windows Tracker
34.4
mMOTA
No paper